Re: Updated GPL release readme
Posted: 13 Jun 2008, 12:24
As the Resurrection Project can never use that exception (ok, we could, if we would manage to get the ok from everyone who ever contributed a significant amount of code), at least this project will never shut its doors.
I guess what Buginator means is that i.e. when Chojun writes his VisiRaptor rendering engine and integrates it into Warzone, he will rip out the original renderer (or it will at least not be comparable to the new one). If he would now turn mad, or vanish from the stage, there would be a significant amount of code which drives WZ missing, only accessible in a closed-source blob. Which would in turn mean that development on his version of Warzone would stop, unless he would have given the sourcecode plus license to someone else before biting the grass (which he might not do at all if he would really turn mad).
You could still compile the current code and run the current version of the game ("... redistributable, linkable ...") but you were unable to improve the portions of the code which do not belong to Warzone anymore. In case of i.e. VisiRaptor you would be stuck with rewriting that library from ground up, if Chojun writes the interfaces nicely (which he no doubt will). If his integration or documentation would be bad, you could as well start fresh from the 2004 version of the code, I guess.
Theoretically you could this way close-source (nearly) the whole game, by rewriting it piece by piece and putting the new code into closed-source libraries. You would only leave as much code in WZ so you could continue to use the name "Warzone 2100". Though: Will this ever happen?
This is a speculation of what might be if ..., a dark utopia, the bad dream of Warzone 2200 fans, not a fact.
So for everyone interested in Warzone 2200 it is all about trust. Trust the developers that they will continue the game, or will pass their proprietary code on to a new generation in case they leave. If you could not trust the Warzone 2200 project, you would be f*ed with that exception.
And even then, there is still the Resurrection Project you could work on. (Or maybe that other fork Rman talked about.)
I don't have much fear here, but then: I am in a special position...
This is somewhat like the KDE - Trolltech relationship of the early days. If Trolltech would have bitten the grass, the KDE project would have been stuck for life with using an old non-developed version of the Qt libraries. Which is why they were so interested in an if-we-die-we-pass-on-the-sourcecode deal with Trolltech. (Now Qt is opensource on every of their platforms, so the problem of Trolltech going broke is less of an issue today.)
To my opinion, writing down any other kind of restrictions for the libraries in the exception would have taken days or weeks of well-payed lawyer-time, and the outcome would have been far more complicated and less understandable than "accessible, redistributable and linkable free of charge". If it would be at all possible. And it would only have restricted the Warzone 2200 project, without any effect on the Resurrection project. (Which is why no one of the involved Resurrection members cared a lot about it.)
Given the time constraints we were under, I think that despite all differences in philosophy, organisation, etc, we worked very well together and created a nice document. (Maybe we can continue that on other fronts as well? )
I am confident that the bad effect the linking exception might have, is pretty much non-existant. (Don't prove me wrong. )
This is especially true for the Resurrection project: The fact that our code does not carry the linking exception, makes it impossible for everyone using the exception, to import our code (without asking the author for granting that exception).
Sheesh, this grew quite a bit over time. (Did I beat kage yet? ) I edited and enhanced it several times, so be sure to read it again even if you think you read it already.
I guess what Buginator means is that i.e. when Chojun writes his VisiRaptor rendering engine and integrates it into Warzone, he will rip out the original renderer (or it will at least not be comparable to the new one). If he would now turn mad, or vanish from the stage, there would be a significant amount of code which drives WZ missing, only accessible in a closed-source blob. Which would in turn mean that development on his version of Warzone would stop, unless he would have given the sourcecode plus license to someone else before biting the grass (which he might not do at all if he would really turn mad).
You could still compile the current code and run the current version of the game ("... redistributable, linkable ...") but you were unable to improve the portions of the code which do not belong to Warzone anymore. In case of i.e. VisiRaptor you would be stuck with rewriting that library from ground up, if Chojun writes the interfaces nicely (which he no doubt will). If his integration or documentation would be bad, you could as well start fresh from the 2004 version of the code, I guess.
Theoretically you could this way close-source (nearly) the whole game, by rewriting it piece by piece and putting the new code into closed-source libraries. You would only leave as much code in WZ so you could continue to use the name "Warzone 2100". Though: Will this ever happen?
This is a speculation of what might be if ..., a dark utopia, the bad dream of Warzone 2200 fans, not a fact.
So for everyone interested in Warzone 2200 it is all about trust. Trust the developers that they will continue the game, or will pass their proprietary code on to a new generation in case they leave. If you could not trust the Warzone 2200 project, you would be f*ed with that exception.
And even then, there is still the Resurrection Project you could work on. (Or maybe that other fork Rman talked about.)
I don't have much fear here, but then: I am in a special position...
This is somewhat like the KDE - Trolltech relationship of the early days. If Trolltech would have bitten the grass, the KDE project would have been stuck for life with using an old non-developed version of the Qt libraries. Which is why they were so interested in an if-we-die-we-pass-on-the-sourcecode deal with Trolltech. (Now Qt is opensource on every of their platforms, so the problem of Trolltech going broke is less of an issue today.)
To my opinion, writing down any other kind of restrictions for the libraries in the exception would have taken days or weeks of well-payed lawyer-time, and the outcome would have been far more complicated and less understandable than "accessible, redistributable and linkable free of charge". If it would be at all possible. And it would only have restricted the Warzone 2200 project, without any effect on the Resurrection project. (Which is why no one of the involved Resurrection members cared a lot about it.)
Given the time constraints we were under, I think that despite all differences in philosophy, organisation, etc, we worked very well together and created a nice document. (Maybe we can continue that on other fronts as well? )
I am confident that the bad effect the linking exception might have, is pretty much non-existant. (Don't prove me wrong. )
This is especially true for the Resurrection project: The fact that our code does not carry the linking exception, makes it impossible for everyone using the exception, to import our code (without asking the author for granting that exception).
Sheesh, this grew quite a bit over time. (Did I beat kage yet? ) I edited and enhanced it several times, so be sure to read it again even if you think you read it already.